The post by F111D (re: Andy Pasztor article) does support those of us who favour the slush/icing theory for the crash.

It is a theory that goes a long way to explain the following:

* near simultaneous loss of power in both engines. (a critical factor that all theories need to explain and a severe limitation on most.)
* slight differences in conditions in each fuel tank can explain the slight difference (seconds) in response.
* why the engines responded properly to start with then dropped down to a thrust level somewhat above flight idle shortly afterwards.
* fuel pump cavitation (from slush retarding fuel suction)
* minimal remaining evidence (remaining slush would melt)
* if slush was mostly sucked out of the system prior to the crash, it would also explain why the fuel would test within specifications. (Jet A-1 Chinese equivalent Jet Fuel #3) It probably wouldn't have taken much water content in the fuel to cause problems if it was in the right location. All (or most) of the water would have frozen out at the temperatures in the fuel tanks and migrated to where it could cause grief.

I don't see any notes in the last AAIB update about EICAS warnings (fuel pressure pump or fuel filter clogged) as noted by CONF iture, although it isn't clear that they would invalidate the theory if they didn't (or did) occur. The AAIB report notes that the high pressure fuel filters were clean after the incident.

It also doesn't rule out there being other factors involved, but it explains events reasonably well without them.

I think that the reason the next update is taking so long is that the incident has a number of subtle causes. AAIB needs to make very sure they have it right. If this theory is correct or mostly correct it could cause huge changes in procedures in long haul flights in extreme cold weather as well as have the potential for other operational or aircraft design changes. If the cause was simple and straightforward, the AAIB would most likely have let us know already.

Quoting AAIB Special Bulletin 1/2008:
 

"- Both engine low pressure fuel filters were clean.
- The fuel oil heat exchangers (FOHE) in both engines were free of blockage.
- The right FOHE was clear of any debris, however the left engine FOHE had some small items of debris on its fuel inlet bulkhead.
- The high pressure filters were clean.

- A sump sample taken from the left and right main fuel tanks shortly after the accident revealed no significant quantities of water."

Although the ice/slush theory is plausible, wouldn't there have been an increased level of water content found in the fuel lines near the fuel filters, fuel oil heat exchangers or servo fuel lines? And if there was a presence of ice/slush, wouldn't some of it have been collected as water in the main tank sumps? Some of the ice/slush, being a solid and heavier than fuel, would likely have found its way to the lowest points in the tanks.

The Andy Pasztor article reminds me of the "industry insider" who claimed for example on jan. 30th [quote] "the fuel crossfeed valve switches were found in the open position and only one valve was open" [unquote]. The AAIB report, published on feb. 18th, explicitly mentioned that both valves indicated that they were closed and they had not been operated during the flight.



Green-dot

Need not be H20 that caused problem
 

Many posts comment on ice/slush as being the primary cause of fuel degradation but this need not be so. The fuel is complex and a prolonged cold-soaking, in a moderately high frequency environment, might produce physical amalgams and/or emulsion of hydrocarbon + H20 and any other material included in the fuel. Give these oddities a few more hours to develop and they may change their physical characteristics again.

Put those oddities back into a temperate climate (Heathrow ground level) and they might disappear altogether, leaving no clue to there former presence.

Difficult to investigate. And, if the AAIB cannot find another cause it might now have to show that the fuel was, or was not, the cause of the accident. I don't think it can leave doubt.


regards, Tanimbar

Xfeeds operated or not? what says FDR?
 

#923 Greendot
....There was a strange re-edit of a lot of posts which became redated 18Feb, so I cant find full versions of the original relevant posts now.
....It had been stated that a "The Captain now recollected operating X feeds" (Plural),with the implication that it was after the engines failed to respond. It was also stated that (I personally cant remember which )
....Either they were set to, open and one was not fully open-
....Or they were set to closed and one was not fully closed.
.... In any case I read that one xfeed finished in a part way position...............So???

Wilyflier

That information didn't come from the AAIB; it was on another aviation site, and it was an 'informed sources have said' kind of report.

It seemed to contradict the subsequent interim report from the AAIB in several ways, so I think it just served to confuse the issue.

No reason to believe that the pilots opened x-feeds, per the AAIB.

Mat Tonkang:
This is off-topic, but just to confirm - the AAIB's work at Lockerbie was purely technical, determining the break-up sequence from the point the bomb went off. It had no jurisdiction on the who, only the how.

Back on topic, the AAIB *is* a branch of the Civil Service, but it is a very different beast to the NTSB. What they do have in common is that they are both politically neutral, so if EMI had not been ruled out, or at least downgraded, they would have mentioned it as a possible contributing factor in the interim reports regardless of where it came from. Their remit is to figure out what happened - nothing more - and by all accounts they are extremely good at it and are world-respected as a result.

To wilyflier #926:

The AAIB special bulletin 1/2008 does not mention any recollection of the crew switching the crossfeed valves or finding one of them open or partially open.

Quoting 1/2008, page 2:

"The total fuel on board was indicating 10,500 kg, which was distributed almost equally between the left and right main fuel tanks, with a minor imbalance of about 300 kg. The crossfeed valves indicated that they were closed and they had not been operated during the flight."

This implies the AAIB has been able to confirm this from the FDR data (and likely from crew interviews) or they would not have made the statement: ". . .and they had not been operated during the flight."

Those are the officially published facts sofar regarding the crossfeed valves. No more, no less.

In light of only 300 kg imbalance between left and right main tank, it is also ruled out that one of the crossfeed valves could have been (partially) open undetected.



Green-dot

I'm not sure about that Green-dot, if crossfeed is open but fuel pumps are not switched off on one side or the other, each wing tank will still feed its own side engine unless fuel pressure is somehow stronger on one side.

If I'm correct one crossfeed will automatically open under a certain type of electrical failure ... !?



... You mentioned that, I mentioned that, it would be nice to know exactly when the APU flap started to move ?

To CONF iture:

You are correct, with all boost pumps running in both main tanks there would be no crossfeed if a crossfeed valve were to be in the open position. However, if only the APU is running (with boost pumps off on the overhead panel) there would be a crossfeed. But that would be a temporary condition.

An undetected open position can be ruled out. First of all, an open crossfeed valve in disagree with the pushbutton would generate a valve light in the pushbutton and EICAS advisory and status messages would be displayed (after 10 seconds). No way it would go unnoticed by the crew.

I am not aware of any automatic crossfeed valve selection. To my knowledge crossfeed valve operation is a manual selection only.

With regards to when the APU flap (assuming you mean the APU door) started to move. That may be a tougher nut to crack. I can think of a few possible scenarios:

- APU started by the crew: Not mentioned by the AAIB in their report S1/2008. Could very well have been the case even if not mentioned by the AAIB.

- APU autostart: For that to happen, power would have to have been removed from left and right transfer buses in flight. That did not happen because the ram air turbine (RAT) would have automatically extended during an APU autostart. There is photographic evidence the RAT did not extend in flight. It extended after impact on the ground.

- APU door open due to other causal factors: Answer to this will have to wait for the final report. This is where several holes in the swiss cheese would have had to line up before a failure of this nature could occur.
As discussed before, i can think of a possible (but very unlikely) scenario where the APU fuel shut off valve control relay could have been affected. The same scenario would have affected the spar valve control relays. Temporary loss of power to the 28v dc hot battery bus, making both the spar valve- and APU fuel s/o vlv control relays susceptible to EMI reverse logic. If (i stress if) EMI was present it could have caused the spar valve control relays to close and the APU fuel s/o vlv relay to open, possibly latched in that position when power was restored until the EMI source disappeared, returning the system to a normal logic condition. The effects would be:

* Closing (or partially closing) of the spar valves restricts fuel flow to the engines.

* Opening the APU fuel s/o vlv control relay, instructing the APU fuel shutoff valve to open, may have triggered a signal to ELMS to open the APU door?

Ready to be flamed . . . .but . . . "when you have eliminated the impossible, whatever remains, however improbable, must be the truth", Sherlock Holmes used to say.


Green-dot

greendot:

How?

Please provide a plausible scenario to drive the relatively high current DC components involved, via spurious radio signals, keep them engaged, and then release them when the radio signal is removed. I won't ask you to explain the presence or strength of said signal, merely to describe the electrical activity of the aircraft systems responding, esp why other systems were apparently completely unaffected, only the fuel control.

I believe it is time to take the strop to Occam's Razor.

http://img117.imageshack.us/img117/8...steninggz4.gif

http://img329.imageshack.us/img329/3...rd01lx1.th.jpg
http://www.airmech.co.uk/forums/showthread.php?t=3650


This gives a good overview and history. CAUTION: link to PDF file:
http://www3.verticalgateway.com/port...s/hirfnprm.pdf

Link to FAA Final rule on HIRF certification as of Feb 2008:
http://rgl.faa.gov/Regulatory_and_Gu...E?OpenDocument

OK, just assuming slushy fuel was the cause, one would think that tests would be carried out with a jet engine and fuel that had been purposely cooled to a state of slushiness. Anyone head of any such tests being performed?

soem dood:

Thanks for the links.

That is what i wondered years ago as well, at a time when we were confronted with uncommanded closing of main fuel shutoff valves (MFSOV), also a 28v dc system but on a different aircraft type. At the time only the valve was affected, no other 28 v dc system components. Regarding a scenario: EMI does not necessarily imply an external source, in this case it was generated in the connector itself.

Something to add to the history:
http://www.corrdefense.org/CorrDefen...df/feature.pdf

and:
http://www.dtic.mil/ndia/2001systems/horne.pdf



With regards to G-YMMM: without the data it remains a theory, nothing more. I do not intend to twist facts to suit theories, it is beter to twist theories to suit facts.

But since i have had first hand experience with a similar issue (as the above links explain), first time it happened it was also thought an impossible scenario, but happen it did.



Green-dot

So it could well be an only Airbus thing, which does not apply to 777, when all AC bus fail, one single fuel pump remains powered and therefore crossfeed valve automatically opens ?

Regarding the ice build-up fuel scenario, I’m very septic that the same conditions develop at the almost exact same time in two separate fuel tanks, especially when one of them had almost 400 more liters than the other one.

I don’t know the oil cooler / fuel warmer architecture, but I find it surprising that the restriction occurs into the very spot where fuel should get warmer ?

Also, still no official mention if thrust has been applied after Top Of Descent but before the event and how both engines did react ?

Greendot, in line with your point that increased susceptibility can come from sources such as degradation of connector internal contact integrity, here is an interesting prior airworthiness directive on the 777:

RIN 2120-AA64 - prelim
http://www.thefederalregister.com/d....12-08-03-30338

RIN 2120-AA64 - final
http://rgl.faa.gov/Regulatory_and_Gu...3?OpenDocument

Thanks for the back story and discussion. I am nothing more than an armchair observer when it comes to air accident investigation, but I have much interest, as I am in a not-too distant line. For those that want an instant perfect answer, and want it now, I will just offer that the NTSB and/or AAIB do good work, and the truth will out, but good original work takes time. One of the most difficult cases ever was on the Boeing "soda can" rudder activation valve that brought down US Air flt 427, and it took a lot of investigation, and frankly, some related incidents (Eastwind 517, United 585) before it was figured out, but they got there. For those who have their interest whetted for following air investigations, but can't wait for the AAIB report, let me suggest any of the following as good reading, from my own bookshelf:

Normal Accidents, Perrow
The Limits of Safety, Sagan
Inviting Disaster, Chiles
Inside the Sky, Langewiesche
Mystery of Flight 427, Adair
Aircraft Investigation Analysis: Final Reports, Walters
Aircraft Safety, Krause
The Logic of Failure, Dörner
Mayday!, Wadley
Human Error, Reason
Columbia Accident Investigation Board, NASA/GPO
Safety in the Skies, Sarsfield

Fiction:
Airframe, Crichton

One of the links extolling the virtues of CPCs noted this about the F16 connectors:

Huge "no-no" using dissimilar metals in an electrical connection that you want to remain reliable. Galvanic action, and other bad things can happen. Pretty surprised to find that all contacts were not initially gold-gold on such a critical application.

As to CPC, the problem with many corrosion preventing compounds is degradation over time. It is hard to beat a good gold to gold mechanical contact. I remember we evaluated some stuff in the eighties call "Tweek" or "Stabilant 22" that was supposedly a 'contact enhancer but it was not proven to improve anything, and I notice in your other presentation, it is specifically not recommended... the water displacement properties of the CPC may be it's best feature, assuming a gas tight connection subsequently...

soem dood,


From one of the links you provided:
It seems dissimilar metals in electrical connections is/was used elsewhere as well.

Interesting book list.
Some of those are on my booklist as well. I am currently reading "Inside The Sky."

Some other noteworthy books on the subject:

Human Factors In Flight, Hawkins
Blind Trust, Nance
The Safe Airline, Ramsden
Pilot Error, Hurst

P.S. The above mentioned book titles should not be percieved as to suggest any link to the cause of this accident. They don't and are mentioned only because they explain the many factors related to accidents in general.


Green-dot

That that shall not be spoken
 

Green dot.
....Speak up in plain language then ,stop hinting.What is your inference that I and the others dont like to say?

Possible and not uncommon reason for the open APU inlet door:
I have seen it happen that the APU ilet door actuator was inoperative. In order to be able to provide APU power if neccessary, the aircraft may opearate under an MEL, with the door secured in its open state throughout the whole operation and an extra percentage of the tripfuel accounted for the aerodynamic loss. So open APU inlet does not neccessarily mean a running APU.

We were at the sim today ( 777 ), finished a bit ahead of time and figured, why not put ourselves on 27L at Heathrow, 160 kt til 4 DME, full flaps and throttle back the engines one by one at 700 ft to try to simulate what happened to BA 038.
The surprise and stress element was of course totally absent.
I disconnected the A/P, went under the G/P and ended doing a hard landing, actually more like a low level stall, just before the runway.
Very interesting experience.
I can't believe people could imagine they could / should have done something different. I was prepared and it just happened so fast I couldn't believe it.
Couple of miles back would have been very tragic.

Latest update pulished today 30 April
http://www.aaib.dft.gov.uk/cms_resou...8%20G-YMMM.pdf